Moving away from EIPs for ENS standards

Historically, ENS has used EIPs as our standardisation process; ENS itself is defined in EIP 137. While this made sense in the past, with the increasing pace of innovation in ENS, and the increasingly bureaucratic process of EIPs, I think the time has come to reconsider this position.

We’ve got a number of protocol changes underway - most significantly, wildcard support and L2 support - and I’d like to shift those over to our own internal standards. As part of the shift, we can copy over existing ENS EIPs and renumber them appropriately, then start adding to them from there.

Thoughts? Objections?

2 Likes

Completely agree. ENS is going to grow in complexity quickly so we should probably adopt a more nuanced improvement proposal process. My first thought was a creating a standard model following Ethereum’s current format or similar. Or maybe it would make more sense to take a more “blockchain agnostic” approach.

2 Likes

I am in general supportive of this proposal. The only downside I can think of is that we will probably lose some of inputs/comments from the existing EIP editors/contributors and it will take some maintenance overhead maintaining the process and would prefer someone apart from @nick.eth to kickstart the initiative given you are already in million different things, so comes down to see if we have someone equivalent of Tim Beiko or Hudson.

(or maybe perfect for @AvsA ? )

I’ve created a category in our docs and moved all the relevant EIPs in there, renumbering them as ENSIPs. I’ve also taken the liberty of finalising a number of them, as they’re widely used despite being drafts in the EIP process.

My next priority now is getting ENSIP-10 (formerly EIP-2544, wildcard resolution) standardised and implemented, as it’s an essential building-block for our L2 solution.

1 Like

How about these?

https://eips.ethereum.org/EIPS/eip-3668
https://eips.ethereum.org/EIPS/eip-2304

Plus Avatar text record specification (somewhere in your gist I always struggle to find where it is).

Also searching EIP site with “ENS” hit a lot more entries https://github.com/ethereum/EIPs/search?q=ENS. Do we move all of them?

1 Like

EIP-2304 is ENSIP-9.

EIP-3668 should remain an EIP because it’s not specific to ENS - but we should probably make an ENSIP for how it can be used in ENS.

I excluded a couple of ENS EIPs that were obsolete or abandoned, but if there are others I have missed we should definitely add them.